From c39ffeff7d1e95aa6c413919fd58633b4995dd68 Mon Sep 17 00:00:00 2001 From: YuryYakhno <52328128+YuryYakhno@users.noreply.github.com> Date: Thu, 12 Oct 2023 13:26:12 +0300 Subject: [PATCH 1/2] Add SantaCoder model in config templates --- src/configTemplates.ts | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) diff --git a/src/configTemplates.ts b/src/configTemplates.ts index 057f1a53..b646cd81 100644 --- a/src/configTemplates.ts +++ b/src/configTemplates.ts @@ -1,4 +1,4 @@ -const templateKeys = ["bigcode/starcoder", "codellama/CodeLlama-13b-hf", "Phind/Phind-CodeLlama-34B-v2", "WizardLM/WizardCoder-Python-34B-V1.0", "Custom"] as const; +const templateKeys = ["bigcode/starcoder", "bigcode/santacoder", "codellama/CodeLlama-13b-hf", "Phind/Phind-CodeLlama-34B-v2", "WizardLM/WizardCoder-Python-34B-V1.0", "Custom"] as const; export type TemplateKey = typeof templateKeys[number]; export interface TokenizerPathConfig { @@ -40,6 +40,20 @@ const StarCoderConfig: Config = { } } +const SantaCoderConfig: Config = { + modelIdOrEndpoint: "bigcode/santacoder", + "fillInTheMiddle.enabled": true, + "fillInTheMiddle.prefix": "", + "fillInTheMiddle.middle": "", + "fillInTheMiddle.suffix": "", + temperature: 0.2, + contextWindow: 2048, + tokensToClear: ["<|endoftext|>"], + tokenizer: { + repository: "bigcode/santacoder", + } +} + const CodeLlama13BConfig: Config = { modelIdOrEndpoint: "codellama/CodeLlama-13b-hf", "fillInTheMiddle.enabled": true, @@ -69,7 +83,8 @@ const WizardCoderPython34Bv1Config: Config = { export const templates: Partial> = { "bigcode/starcoder": StarCoderConfig, + "bigcode/santacoder": SantaCoderConfig, "codellama/CodeLlama-13b-hf": CodeLlama13BConfig, "Phind/Phind-CodeLlama-34B-v2": PhindCodeLlama34Bv2Config, "WizardLM/WizardCoder-Python-34B-V1.0": WizardCoderPython34Bv1Config, -} \ No newline at end of file +} From 51da0e5f780f5d68a2bc5e49524c5329910ab9bf Mon Sep 17 00:00:00 2001 From: YuryYakhno <52328128+YuryYakhno@users.noreply.github.com> Date: Thu, 12 Oct 2023 14:22:41 +0300 Subject: [PATCH 2/2] Add SantaCoder to package config --- package.json | 1 + 1 file changed, 1 insertion(+) diff --git a/package.json b/package.json index c7f9aafd..a4c92fb0 100644 --- a/package.json +++ b/package.json @@ -81,6 +81,7 @@ "type": "string", "enum": [ "bigcode/starcoder", + "bigcode/santacoder", "codellama/CodeLlama-13b-hf", "Phind/Phind-CodeLlama-34B-v2", "WizardLM/WizardCoder-Python-34B-V1.0",