diff --git a/codellama-7b-instruct.yaml b/codellama-7b-instruct.yaml index 6bafb2fd..4208fa8c 100644 --- a/codellama-7b-instruct.yaml +++ b/codellama-7b-instruct.yaml @@ -9,12 +9,13 @@ urls: - https://huggingface.co/TheBloke/CodeLlama-7B-Instruct-GGUF config_file: | + backend: llama + context_size: 4096 parameters: model: codellama-7b-instruct.Q4_K_M.gguf top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 4096 template: chat_message: llama2-chat-message diff --git a/gorilla.yaml b/gorilla.yaml index b7a2d053..d9b25580 100644 --- a/gorilla.yaml +++ b/gorilla.yaml @@ -8,13 +8,13 @@ license: "N/A" config_file: | backend: llama + context_size: 1024 + f16: true parameters: model: gorilla top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 1024 - f16: true template: completion: gorilla chat: gorilla diff --git a/gpt4all-j-groovy.yaml b/gpt4all-j-groovy.yaml index 08630aee..b7f5b191 100644 --- a/gpt4all-j-groovy.yaml +++ b/gpt4all-j-groovy.yaml @@ -6,12 +6,12 @@ urls: - https://gpt4all.io config_file: | backend: gpt4all-j + context_size: 1024 parameters: model: ggml-gpt4all-j-v1.3-groovy.bin top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 1024 template: completion: "gpt4all-completion" chat: gpt4all-chat @@ -33,4 +33,4 @@ prompt_templates: The prompt below is a question to answer, a task to complete, or a conversation to respond to; decide which and write an appropriate response. ### Prompt: {{.Input}} - ### Response: \ No newline at end of file + ### Response: diff --git a/gpt4all-j.yaml b/gpt4all-j.yaml index 6b7830d1..ab9bac1d 100644 --- a/gpt4all-j.yaml +++ b/gpt4all-j.yaml @@ -6,12 +6,12 @@ urls: - https://gpt4all.io config_file: | backend: gpt4all-j + context_size: 1024 parameters: model: ggml-gpt4all-j.bin top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 1024 template: completion: "gpt4all-completion" chat: gpt4all-chat @@ -33,4 +33,4 @@ prompt_templates: The prompt below is a question to answer, a task to complete, or a conversation to respond to; decide which and write an appropriate response. ### Prompt: {{.Input}} - ### Response: \ No newline at end of file + ### Response: diff --git a/gpt4all-l13b-snoozy.yaml b/gpt4all-l13b-snoozy.yaml index d25954d7..d037a1bf 100644 --- a/gpt4all-l13b-snoozy.yaml +++ b/gpt4all-l13b-snoozy.yaml @@ -6,12 +6,12 @@ urls: - https://gpt4all.io config_file: | backend: gpt4all-llama + context_size: 1024 parameters: model: ggml-gpt4all-j-v1.3-groovy.bin top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 1024 template: completion: "gpt4all-completion" chat: gpt4all-chat @@ -33,4 +33,4 @@ prompt_templates: The prompt below is a question to answer, a task to complete, or a conversation to respond to; decide which and write an appropriate response. ### Prompt: {{.Input}} - ### Response: \ No newline at end of file + ### Response: diff --git a/guanaco.yaml b/guanaco.yaml index dd5abe63..3eb58045 100644 --- a/guanaco.yaml +++ b/guanaco.yaml @@ -8,12 +8,12 @@ urls: config_file: | backend: llama + context_size: 1024 parameters: model: guanaco top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 1024 template: completion: guanaco-completion chat: guanaco-chat @@ -34,4 +34,4 @@ prompt_templates: content: | ### Instruction: {{.Input}} - ### Response: \ No newline at end of file + ### Response: diff --git a/hippogriff.yaml b/hippogriff.yaml index 7e678e67..6ab9a7f4 100644 --- a/hippogriff.yaml +++ b/hippogriff.yaml @@ -7,18 +7,18 @@ license: "N/A" config_file: | backend: llama + context_size: 1024 + f16: true + mmap: true parameters: model: hippogriff top_k: 40 temperature: 0.1 top_p: 0.95 - context_size: 1024 roles: user: "USER:" system: "SYSTEM:" assistant: "ASSISTANT:" - f16: true - mmap: true template: completion: hippogriff-completion chat: hippogriff-chat @@ -30,4 +30,4 @@ prompt_templates: ASSISTANT: - name: "hippogriff-completion" content: | - {{.Input}} \ No newline at end of file + {{.Input}} diff --git a/hypermantis.yaml b/hypermantis.yaml index 33fc9a09..35aa7d2c 100644 --- a/hypermantis.yaml +++ b/hypermantis.yaml @@ -9,12 +9,12 @@ license: "" config_file: | backend: llama + context_size: 2048 parameters: model: hypermantis top_k: 80 temperature: 0.2 top_p: 0.7 - context_size: 2048 template: completion: hypermantis-completion chat: hypermantis-chat @@ -38,4 +38,4 @@ prompt_templates: ### Response: - ASSISTANT: \ No newline at end of file + ASSISTANT: